Setting Up the Loudspeaker Locations with Two Pieces of String

Many times a trigonometry book is not at hand, and you have to set up a surround monitor system. Here is how:

• Determine the distance from each of the left and right loudspeakers to the principal listening position to be used. Home listening is typically done at 10ft (3m), but professional listening is over a wide range due to differing requirements from a large scoring stage control room to a minimum sized booth in a location truck.

• Cut a piece of string to the length of the listening distance, or use a marked XLR extension, and make the distance between the left and right loudspeakers equal to the distance between each of them and the prime listening location. This sets up an equilateral triangle, with a 60° subtended angle between left and right loudspeakers.

• Place the center speaker on the centerline between left and right speakers. Use the full string length again to set the distance from the

listener to the center speaker, putting the front three loudspeakers on an arc, unless an electronic time delay compensation is available (discussed below).

• Use two strings equal in length to the listening distance. Place one from the listener to the left loudspeaker, and the other perpendicular to the first from the listener's location and to the outside of the front loudspeakers. Temporarily place a surround loudspeaker at this location, which is 90° from the left loudspeaker; 90° plus the 30° that the left is from center makes 120°. This angle is within the tolerance of the standard, but to get it right on, swing the loudspeaker along an arc towards the front by one-third of the distance between the left and center loudspeaker. This places it 110° from center front, assuming all the loudspeakers are at a constant distance from the listener.

• Repeat in a mirror image for the right surround channel (Fig. 2-5).

Fig. 2-5 The speaker layout from AES TD1001 and ITU Recommendation 775.

Setup Compromises

Frequently, equipment, windows, or doors are just where the loudspeakers need to be placed. Following are compromises that can be made if necessary:

• Generally, in front hearing is about three times less sensitive to errors in elevation than to horizontal errors, while at the sides and back these listening errors become much larger and therefore freedom in placement is greater. Therefore, it is permissible to elevate loudspeakers above obstructions if necessary. It is best to elevate the loudspeakers only enough to clear obstructions, since if they are too high, strong reflections of sound will occur off control surfaces at levels that have been found to be audible.

• The tolerance on surround loudspeaker placement angles is wide. Probably both surround loudspeakers should use close to the same angle from center, although the range is ±10° from the ±110° angle. Also, surround loudspeaker placement is often a compromise, because if the producer sits behind the engineer, then the surround angles for the producer are significantly less than for the engineer, especially in smaller rooms. With this in mind, it may be useful to use a somewhat wider angle from the engineer's seat than 110° to get the producer to lie within the surrounds, instead of behind them.

• With a sufficiently low crossover frequency and steep filter, along with low distortion from the subwoofer, placement of the subwoofer(s) becomes non-critical for localization. Thus, it or they may be placed where the smoothes! response is achieved. This often winds up in a front corner, or if two are used, one in a front corner and one halfway down a side wall, or centered along the two side walls, to distribute the driving points of the room and "fill in" the standing wave patterns.

Center

In many listening situations it may be impossible to place the center loudspeaker at 0° (straight ahead) and 0° elevation. There is probably other equipment that needs to be placed there in many cases. The choices, when confronted with practical situations involving displays or controls, are:

• above the display/control surfaces,

• below the display/control surfaces,

• behind the display.

Placement above the display/control surface is common in many professional facilities, but such placement can carry a penalty in many

instances the sound splash off the control surface is strongly above audibility. Recommended placement for monitor loudspeakers is to set them up so that the principal listener can see the whole speaker, just over the highest obstruction. This makes the sound diffract over the obstruction, with the obstruction, say a video monitor, providing an acoustic shadow so far as the sound control surface is concerned. The top of the obstruction can be covered in thin absorbing material to absorb the high frequencies that would otherwise reflect off the obstruction.

Another reason not to elevate the center speaker too much is that most professional monitor loudspeakers are built with their drivers along a vertical line, when used with the long dimension of the box vertically. This means that their most uniform coverage will be horizontal, and vertically they will be worse, due to the crossover effects between the drivers. Thus, if a speaker is highly elevated and tipped down to aim at the principal mixer's location, the producer's seat behind the mixer is not well covered—there are likely to be mid-frequency dips in the direct-field frequency response. (The producer's seat also may be up against the back wall in tight spaces, and this leads to more bass there than at the mixer's seat.)

A third reason not to elevate the studio monitor loudspeaker too much is that we listen with a different frequency response versus vertical angle. Called head-related transfer functions (HRTFs), this effect is easily observed. Playing pink noise over a monitor you are facing, tip your head up and down; you will hear a distinct response change. Since most listeners will not have highly elevated speakers, we should use similar angles as the end user in order to get a similar response.

A position below the display/control surface is not usable in most professional applications, for obvious reasons, but it may be useful in screening rooms with direct view or rear-projection monitors. The reason that this position may work is that people tend to locate themselves so they can see the screen.This makes listeners that are further away elevated compared to those closer. The problem with vertical coverage of the loudspeaker then is lessened, because the listeners tend to be in line with each other, viewed from the loudspeaker, occupying a smaller vertical angle, with consequently better frequency response.

When it is possible, the best solution may well be front projection with loudspeakers behind the screen using special perforated screens. Normally perforated motion picture theater screens have too much high-frequency sound attenuation to be useful in video applications, but in the last few years several manufacturers have brought out screens

with much smaller perforations that pass high-frequency sound with near transparency, and that are less visible than the standard perforations. However, perforation patterns may interact with the pixel pattern in modern projectors and produce moire patterns. For such cases, the woven and acoustically transparent screens become a necessity, such as those of Screen Research.

By the way, it is not recommended that the center loudspeaker be placed "off center" to accommodate a video monitor. Instead, change the loudspeaker elevation, raising it above the monitor, and moving it back as shown in the Fig. 2-6, creating an "acoustic shadow" so the effect of the direct reflection off the console is reduced.This is permissible since listening is more sensitive to errors in the horizontal plane than in the vertical one.

A "top and bottom" approach to the center channel to try to center a vertical phantom image on the screen may have some initial appeal, but it has two major problems. The first is that at your eardrum these two have different responses, because of the different HRTFs at the two angles. The second is that then you become supremely sensitive to seated ear height, as the change of an inch off the optimum could bring about an audible comb filter notch at high frequencies.

Left and Right

One problem that occurs when using sound accompanying a picture with the ±30° angle for left and right speakers is that they are unlikely to be placed inside the picture image area using such a wide angle. Film sound relies on speakers just inside the left and right extremities of the screen to make front stereo images that fit within the boundaries of screen. The reason for this is so that left sound images match left picture images and so forth across the front sound field. But when film is translated to video, and the loudspeakers are outside instead of inside the boundaries of the screen, some problems can arise. Professional listeners notice displacement between picture and sound images in the horizontal plane of about 4°, and 50% of the public becomes annoyed when the displacement reaches 15°. So, in instances where the picture is much smaller than 60°, there may need to be a compromise between what is best for sound (±30°), and what is best for sound plus picture (not much wider than 4° outside the limits of the picture). For those mixing program content with no picture, this is no consideration, and there are programs even with a picture where picture-sound placement is not so important as it is with many movies. Thus for a 32° total horizontal subtended angle

Fig. 2-6 When an associated video picture or a computer monitor are needed, the alternatives for

center are above, below, or behind the associated picture, each of which has pros and cons.

high-definition screen,³ speakers at ±30° are seen as too wide, and they would normally be brought in to be just outside the picture (or just within it if front projection and perforated screens are in use). This does limit auditory source width of direct channels and so is a compromise for sound, but is necessary for conditions where picture and sound must match localization.

Surround

Even if the surround loudspeakers are the same model as the fronts, there will still be a perceived frequency response difference. This is due to the HRTFs, the fact that the frequency response in your ear canal determines the spectrum that you hear, not the frequency response measured with a microphone. Your head has a different response for sound originating in front compared to the rear quadrant, and even when the sound fields are perfectly matched at the position of the head, they will sound different. A figure in Chapter 6 shows the frequency response difference in terms of what equalization has to be applied to the surround loudspeaker to get it to match spectrum with the center front. (This response considers only the direct sound field, and not the effects of reflections and reverberation, so practical situations may differ.)

The surrounds may be elevated compared to the fronts without causing much trouble. As they get more overhead, however, they may become less distinguishable from one another and thus more like a single monophonic channel, so too high an angle is not desirable. Experiments into surround height reveal little difference from 0° elevation to 45° elevation for most program material. Some mixers complain, however, about elevated monitors if the program contains audience sound like applause: they don't like the effect that the listeners seem located below the audience in these cases.

Subwoofer

In one case, for FCC mandated listening tests to low-bit-rate codecs for Digital Television, I first set up two subwoofers in between the left and center, and center and right loudspeakers. Since we had no means to adjust the time delay to any of the channels (discussed below), I felt that this would produce the best splice between the main channels and the subwoofer. Unfortunately, these positions of driving the room with the subwoofers, which were set up symmetrically in the room, produced lumpy frequency response. I found that by moving one of

³~^'\^'ls angle is set by noting HDTV has 1920 pixels horizontally, and that 20/20 vision corresponds to an acutance of 1/60° of arc. Dividing 1920 by 60 yields 32°.

the subs one-half way between the front and surround loudspeakers, the response was much smoother. This process is called "placement equalization," and although it requires a spectrum analyzer to do, it is effective in finding placements that work well.

The use of a common bass subwoofer for the 5 channels is based on psychoacoustics. In general, low-frequency sound is difficult to localize, because the long wavelengths of the sound produce little difference at the two ears. Long-wavelength sound flows freely around the head through diffraction, and little level difference between the ears is created. There is a time difference, which is perceptible, but decreasingly so at lower frequencies. This is why most systems employ five limited bandwidth speakers, and one subwoofer doing six jobs, extending the 5 channels to the lowest audible frequencies, and doing the work of the 0.1 channel.The choice of crossover frequency based on finding the most sensitive listener among a group of professionals, then finding the most sensitive program material, and then setting the frequency at two standard deviations below the mean found by experiment, resulted in the choice of 80 Hz for high-quality consumer systems.

Another factor to consider when it comes to localizing subwoofers is the steepness of the filters employed, especially the low-pass filter limiting the amount of mid-range sound that reaches the subwoofer. If this filter is not sufficiently steep, such as 24dB/octave, even if it is set to a low-frequency, higher-frequency components of the sound will come through the filter, albeit attenuated, and still permit the subwoofer to be localized. The subwoofer may also be localized in two accidental ways: through distortion components and through noise of air moving through ports. Both of these have higher-frequency components, outside the band of the subwoofer, and may localize the loudspeaker. Careful design for distortion, and locating the speaker so that port noise is directed away from direct listening path, are helpful.

No one has suggested that each of the 5 channels must be extended down to 20 Hz individually.The problem with doing this in any real room is that the low frequency response would vary dramatically from channel to channel, due to the different driving points of the room. Remember that even "full-range" professional monitors have a cutoff of 40-50 Hz, so using bass management below this frequency is still valuable. With very large PMC monitors at CES some years ago, I found the best splice to theWhise 616 subwoofer (a 5-ft cube with 4 15-in. drivers which had a low-frequency rolloff of -1 dB at 16Hz), to be at 25 Hz! And switching on and off the subwoofer for the 16-25 Hz content was audible on some of the concert hall recorded program material! But the marketplace remains practical, and the best cinema subs usually are down 1 dB at around 24-26Hz.

Setup variations

Дата добавления: 2015-10-30